May 7, 1986 - endocrine and neuronal cells (O'Connor et al., 1983; Lloyd and. Wilson ...... Settleman,J., Nolan,J. and Hogue Angeletti,R. (1985a) J. Biol.
The EMBO Journal vol.5 no.7 pp.1495-1502, 1986
The primary structure of bovine chromogramin A: a representative of a class of acidic secretory proteins common to a variety of peptidergic cells
Ulrich M.Benedum, Patrick A.Baeuerle, David S.Konecki, Rainer Frank', John Powell2, Jacques Mallet2 and Wieland B.Huttner Cell Biology Programme and 1Biochemical Instrumentation Programme, European Molecular Biology Laboratory, Postfach 102209, D-6900 Heidelberg, FRG, and 2Department of Neurobiology, CNRS, F-91190 Gifsur-Yvette, France Communicated by K.Simons
We have determined the primary structure of bovine chromogranin A as a frst step in the elucidation of the function of this widespread protein. After oligonucleotide screening of a cDNA library of bovine adrenal medulla, a clone (insert length 1.9 kb) containing the entire coding region for chromogranin A was isolated and sequenced. The authenticity of the sequence was verified by comparison with N-terminal, several internal, and C-terminal amino acid sequences as well as the amino acid composition of chromogranin A. The cDNA clone hybridized to an mRNA of 2.1 kb and, after in vitro transcription-translation, yielded a polypeptide with a snimlar electrophoretic mobility in SDS gels to chromogranin A . The polypeptide chain of chromogranin A comprises 431 amino acid residues, corresponding to an unmodified protein of 48 kd, and is preceded by a cleaved signal peptide of 18 amino acid residues. Interesting features of the chromogranin A structure include (i) repeated clusters of glutamic acid residues, (ii) the occurrence of eight potential dibasic cleavage sites, six of which are located in the C-terminal domain, and (iii) the presence, in the N-terminal domain, of -Arg-Gly-Asp- (RGD), a three amino acid sequence involved in the binding of several constitutively secreted proteins to cell membranes. Key words: adrenal medulla/Arg-Gly-Asp/cDNA sequence/ neuropeptides/secretogranins Introduction Recent work has shown that secretory granules of a wide variety of endocrine and neuronal cells contain not only their specific regulatory peptides but also members of another class of proteins which are released with regulatory peptides upon stimulation. The first indication came from the work of Cohn et al. (1982) who noted that chromogranin A of bovine adrenal medulla, initially thought to be related to catecholamine secretion (see Winkler, 1976; Winkler and Carmichael, 1982), was very similar to secretory protein I of the bovine parathyroid gland which is secreted with the parathyroid peptide hormone (Cohn and MacGregor, 1981). Subsequently, several investigators reported the presence of chromogranin A in various peptidergic endocrine and neuronal cells (O'Connor et al., 1983; Lloyd and Wilson, 1983; Cohn et al., 1984; Somogyi et al., 1984). Interestingly, the occurrence in many different peptidergic cells is not a feature that is limited to chromogranin A. It was recently reported that secretogranin I (Rosa et al., 1985b; Fischer© IRL Press Limited, Oxford, England
Colbrie et al., 1985; see footnote) and secretogranin H (Rosa and Zanini, 1983; Rosa et al., 1985a, b), two proteins first described as the major sulfated secretory proteins of rat PC12 cells (Lee and Huttner, 1983) and bovine anterior pituitary (Zanini and Rosa, 1981; Rosa and Zanini, 1981), respectively, also occur in secretory granules of a wide variety of peptidergic endocrine and neuronal cells. Secretogranin I, secretogranin II and chromogranin A are clearly distinct by several biochemical and immunological criteria, and their abundance relative to each other varies between different cell types (Rosa et al., 1985b). Nevertheless, the three proteins may be regarded as members of one protein class. They are all very acidic, heat-stable proteins with post-translationally added 0-linked oligosaccharide, phosphate and sulfate moieties and undergo partial proteolytic processing. Furthermore, they are all selectively expressed in peptidergic cells and are sorted like the peptides to secretory granules (see Rosa et al., 1985b). In fact, in endocrine and neuronal cells, chromogranin A and secretogranins I and H are the most widespread targets known for sorting to the regulated pathway of secretion (see Kelly, 1985). We therefore regard these proteins as ideal for identifying common structural components required for the sorting process. The widespread occurrence of chromogranin A and secretogranins I and II suggests important biological roles for these proteins. The observation that these three proteins belong to one class may indicate that their functions are quite similar and related to the common features of these proteins. Proposals for the functions of these proteins include a role (i) in the packaging and sorting of regulatory peptides to the regulated pathway of secretion, (ii) in the organization of the content of secretory granules, (iii) as carrier proteins for regulatory peptides after secretion, (iv) as precursors of regulatory peptides, and (v) as proteins which themselves have some biological activity on target cells (Rosa et al., 1985a, b; Cohn et al., 1984; Somogyi et al., 1984). However, no direct evidence in support of any of these possible functions has been obtained so far. As a step towards elucidating the sorting of chromogranin A and secretogranins I and H as well as the functions of these proteins, we decided to clone cDNAs encoding these proteins and to determine their primary structure. Here we report on the primary structure of chromogranin A.
Results Cloning of chromogranin A cDNA A cDNA library from bovine adrenal medulla was screened with a synthetic 59-mer oligonucleotide (Figure 1). The sequence of this oligonucleotide was derived from the recently reported Nterminal amino acid sequence of a 65-kd fragment of chromogranin A (see footnote) (Settleman et al., 1985b). The oligonucleotide was synthesized without degenerations, choosing the codon most frequently used in mammals (Grantham et al., 1980) where more than one triplet was possible. The reasons for choosing the sequence of the 65-kd fragment included the following. 1495
U.M.Benedum et al. 65 kDa protein
sequence:
MET PRO VAL ASN ILE PRO MET ASN LYS GLY GLU VAL GLU VAL MET LYS ILE ILE VAL GLU
deduced mRNA
sequence:
C A C A A A A A A A AUG CCN GUN AAU AUC CCN AUG AAU AAG GGN GAG GUN GAG GUN AUG AAG AUC AUC GUN GAG U
59-mer oligonucleotide:
U
U
TAC GGG CAC TTG TAG GGG TAC TTG TTC CCG CTC CAC CTC CAC TAC TTC TAG TAG CAC CT
0M ,1'''5''''5
'
'
'
11''ll
HH. H''H,H''
CgA mRNA
sequence:
CUG CCU GUG AAC AGC CCC AUG AAU AAA GGG GAC ACU GAG GUG AUG AAG UGU AUC GUC GAG
CgA protein
sequence:
LEU PRO VAL ASN SER PRO MET ASN LYS GLY ASP THR GLU VAL MET LYS CYS ILE VAL GLU
Fig. 1. Sequence comparison of (first line) the N-terminus of the 65-kd fragment of chromogranin A reported by Settleman et al. (1985b), (second line) the back-translated mRNA, (third line) the complementary 59-mer oligonucleotide synthesized, (fourth line) the actual chromogranin A (CgA) mRNA and (fifth line) the corresponding protein.
First, this sequence is sufficiently homologous to the N-terminal sequence of the intact 75-kd chromogranin A (Hogue Angeletti, 1977) to expect cross-hybridization. Second, in view of the reported divergence between the N-terminal sequences of intact chromogranin A and the 65-kd chromogranin A fragment (Settleman et al., 1985b), one would expect the latter sequence to be an internal sequence of chromogranin A, located at most 10 kd from the N-terminus. This should allow the identification of chromogranin A cDNA clones even if they are not full length. Third, there is the possibility of isolating secretogranin I cDNA clones since the sequence of the 65-kd fragment is sufficiently homologous to the N-terminal sequence of the 100-kd protein of Settleman et al. (1985b) which probably is identical to secretogranin I (Rosa et al., 1985b). Approximately 14 000 colonies of a bovine adrenal medulla cDNA library were screened with the 59-mer oligonucleotide. Five positive clones were obtained. Restriction analysis of these clones showed that the longest insert was -2 kb. This clone (CgA1) was cut with several restriction enzymes. Only one fragment, a PstI-BamHI fragment derived from the 5'-end of the insert (see Figure 2A), was found to hybridize with the 59-mer oligonucleotide. This DNA fragment was isolated, subcloned and sequenced. A 24-amino-acid segment of the deduced protein sequence was found to be identical to the N-terminal sequence of bovine chromogranin A determined by Kruggel et al. (1985) and of the 75-kd chromogranin A reported by Hogue Angeletti (1977) except for two mismatches in positions 2 and 19 where we found proline and valine instead of arginine and arginine, respectively. Sequencing the six N-terminal amino acid residues of chromogranin A from bovine adrenal medulla confirmed the presence of proline in position 2 (data not shown). Moreover, the Nterminal sequence deduced from the cDNA was identical to that of chromogranin A from the bovine parathyroid determined by Cohn et al. (1981) (see footnote). Complete nucleotide and deduced amino acid sequence of chromogranin A The complete nucleotide sequence of the mRNA derived from the cDNA clone CgA1 was obtained as illustrated in Figure 2A and is shown in Figure 2B. It is 1881 bp long [not including the poly(A) tail] and contains a 182 bp untranslated region at the 5'-end, a 1347 bp coding region, and a 352 bp untranslated region at the 3'-end which includes the polyadenylation signal AAUAAA. The open reading frame codes for a 449-amino-acidlong protein of 48 kd. The first 18 amino acid residues constitute the signal sequence responsible for targeting the nascent polypeptide to the membrane of the rough endoplasmic reticulum. The signal sequence ends with an analine residue which is a suitable cleavage site for signal peptidase (see Hortsch and Meyer, 1986). The subsequent sequence is identical to the N-terminal sequence of mature chromogranin A from bovine chromaffin granules (Kruggel et al., 1985; Figure 3, peptide IV). To establish the authenticity of clone CgA1, its deduced amino 1496
Table I. The amino acid composition deduced from the nucleotide sequence (see Figure 2B) is given without the signal peptide Amino acid composition of bovine chromogranin A (residues/100 residues) Deduced from cDNA Protein analysis Adrenal medulla Parathyroid Adrenal medulla Ala Arg Asn Asp Asx Cys Gin Glu Glx Gly His Ile Leu Lys Met Phe Pro Ser Thr Trp Tyr Val
8.1 6.5 2.1 5.1 7.2 0.5 5.3 18.8 24.1 7.7 1.6 1.2 7.4 8.1
1.6 1.2 8.8 8.1 2.1 1.4 1.0 3.5
8.9 6.6
8.5 6.0
7.1 0.5
8.0 0.2
23.2 7.9 1.6 1.0 7.4 8.7 1.6 1.5 8.7
22.5 8.1 1.9 1.4 7.3 8.4 1.4 1.7 9.2 7.7 2.6 1.4 1.0 3.9
7.6 2.1 1.1 1.0 3.5
The amino acid composition data for the chromogranin A protein from the parathyroid and from adrenal medulla chromaffin granules are taken from Cohn et al. (1982).
acid sequence was compared to newly generated peptide sequences of bovine chromogranin A purified from adrenal medulla. First, tryptic peptides of chromogranin A were separated by reverse-phase h.p.l.c. followed by N-terminal amino acid sequencing of selected peaks (Figure 3). All sequences obtained were found at various positions in the open reading frame and all except for the N-terminally located peptide IV start at tryptic cleavage sites (compare Figure 2B). Second, chromogranin A was subjected to simultaneous digestion with carboxypeptidases B and P, followed by analysis of the released amino acids (Figure 4). The time course of the released amino acids supported the C-terminal sequence -Glu-Leu-Arg-Arg-Gly deduced from the cDNA (compare Figure 2B). Table I shows that the deduced amino acid composition for chromogranin A without its signal peptide is in excellent agreement with the reported amino acid compositions of chromogranin A from bovine adrenal medulla and parathyroid (Cohn et al., 1982). Together, the results shown in Figures 3 and 4 and Table I lead us to conclude that the sequence shown in Figure 2B represents the complete amino acid sequence of the bovine chromogranin A precursor.
Molecular cloning of chromogranin A
A 100 bp
-1
-18
MetArgSerAlaAlaValLeuAlaLeuLeuLeuCysAlaGlyGlnValIleAla IV
+1
41
10
20
30
50
60
70
100
110
9_0
81
Z
*
Z
40
**
*
80
120
*
|~~*
1 40 1 60 121 150 1 30 GAGAGAGAUGCAGCGAGGGCCCUAGUCCUCGACAGCAUGGACGCACACGAGACGGGGCAGCCCAACAGAGGAGCAGCUCCGCGCAAGGAGUCGCGGAGGCUGCCAAGCGAUGCGAGCAAGAUGGAGCC
GlLeAaLsGueuhrlGlLyAgLuGulylulGlGlGuGuAprosArgSerMeArglaaLeuSrhArgAelaArgGlyTyrGlyPheArglyProl
200 161 180 170 190 AGCUCAGGUUGCC CAUCGCUCCAGAGGCUCUCCUGCUUCCUGCUCUCCGGGGCUA CAGCUGGCCAAGAGCUGCGCGGCCGAGAACGCGCCCAGAGGGGGAGCGAUGGAGGCGGAGCCCGGCUCG
GlLeuGrVlAnSeuAgrProMetrArnyGProAsnSerGlnGuVaspSetLyCyslGlualaGlyLuValIleeuGlnpThrgGlyTyrLyProSerProMetProValSerLysGluGlysPerAlaAs 220 UCAUCUCUGCC 201 210 CGGGCUGUACAGCUCGCGGCAUGGAAAGGCCGACACUCCAGGAGAGACAGCUGUACUGAGG
241
250
260
III AACCUCCAGGGCC
240 GCCCAGCCCGGAAUGCCAGUCGAAGGAGGUGGGCGUUUGCCAC 230
GGCACUCCGAGGACACUGACGUGGUCCCUCUCAUGUUGCCGACCAUGCAGGAAUUUGCUGAAGAGUCAAGCCGCACCUCGCUCUCCAAGGAGCCAAGGCCGCCGCACACAGCC AGCAAGCCAGCCCACC 281
3 21
361
V
I
*
*
290
300
310
33 0
34 0
35 0
370
380
390
*
1316
136
280
II
270
1196
*
1676
320
36 0
*
*
400
4 10 4 30 420 401 ArgArgProGluAspGlnGluLeuGluSerLeuSerAlaIleGluAlaGluLeuGluLysValAlaHisGlnLeuGluGluLeuArgArgGly.
UACUGACCUUUAGGUGAAAGCUGAGGACUGCUGACUAUAUUGUGUAUGAAGAUUUAUCUAACGGAAAAUAAAUCUGCUCUGGGCA-(Gpoly-A) Fig. 2. Part A. Structure of the chromogranin A cDNA clone CgAl. Only some of the restriction sites are shown. The thick line between the two Nar sites indicates the 1347-bp open reading frame. The solid bar between the Narl and BamHl sites indicates the hybridizing region of the 59-mer oligonucleotide. The three open bars indicate the single-stranded probes used for northern blot analysis (see Figure 6). The arrows show the direction and length of the M 13-clones used for sequencing. Part B. The complete nucleotide and deduced amino acid sequence of the bovine chromogranin A precursor (clone CgA 1). The negative numbers above the amino acid sequence indicate the signal peptide. The lines with Roman numbers above the amino acid sequence indicate the positions of the peptide sequences shown in Figure 3. The C-terminal sequence of Figure 4 is marked with a line. Asterisks indicate dibasic sites. Triangles point to the polyadenylation signal of the miRNA. The shaded bar indicates the RGD sequence.
Structural features of chromogranin A According to the hydropathy plot shown in Figure 5, chromogranin A (without its signal peptide) is predicted to be a very hydrophilic molecule throughout the length of its polypeptide and to lack significant hydrophobic domains. The repeatedly occur-
ring peaks of hydrophilicity often coincide with clusters of glutamic acid residues. The predicted hydrophilicity is in accordance with the known biochemical properties of the protein (see Winkler, 1976; Winkler and Carmichael, 1982), in particular with is property to remain soluble after boiling (Rosa et al., 1497
U.M.Benedum et al.
0.5
W
III
II
I
FMDQLAK GEWAH AVPEEESPP1 -4
A210
o
0
0.5
10
20
30
time (m)n)
40
H" I I~ ~Ei3Iv V S IV,I,,wJ*
100
A~~~~~~~ -rD
0
o~~~~~~~~~~~~~
0~~~~~~~~~~~~~~~~ 40 0 30 20 10
time (min)
50
60
Fig. 3. High performance liquid chromatography of tryptic peptides of chromogranin A. Top. Peptides eluted from the Cl8-cartridge with 30% methanol were chromatographed on an LKB-system equipped with a TSK ODS-120T column, using an acetonitrile gradient from 0 to 30%, as indicated by the dashed line. Bottom. Peptides eluted from the C18-cartridge (after washing with 50% methanol) with 100% methanol were chromatographed on a Beckman-system equipped with an Ultrosphere ODS column, using an acetonitrile gradient from 0 to 100%, as indicated by the dashed line. The indicated peptide peaks were analyzed by N-terminal sequencing in a gas-phase-sequenator. The sequences obtained are given in the single letter amino acid code. The sequence of peptide IV is identical to the previously reported N-terminal sequence of chromogranin A (Kruggel et al, 1985). Top panel, inset. A Coomassie Blue-stained SDS polyacrylamide gel of the soluble proteins of bovine chromaffin granules is shown. Positions of molecular weight markers (97, 67, 45 kd) are indicated by arrowheads. The bracket marks chromogranin A.
1498
Molecular cloning of chromog_nin A
200
1
2 3
-0 c0 ro (A
.
_
100
0
E
0
0
0 3
5
15
45
time (min) Fig. 4. C-terminal sequencing of chromogranin A. The time course of release of amino acids from chromogranin A by simultaneous digestion with carboxypeptidases B and P is shown, confirming the deduced C-terminal sequence of -Glu-Leu-Arg-Arg-Gly.
28S-o
18S
-
-2
L
AS
HPhobic
0
v WY
~~~~~~~~~~~~~~~HPhil1Ic
V2
-18 +1
100
200
300
400
Fig. 5. Hydropathy plot of the chromogranin A precursor according to Kyte and Doolittle (1982). The numbering of amino acid residues is the same as in Figure 2B. HPhobic, hydrophobic; HPhilic, hydrophilic.
1985b). The hydropathy plot lends no support to the existence of two membrane-spanning segments, which one would have to postulate in view of the ectoplasmic location of both the Nterminus (Hogue Angeletti, 1977) and C-terminus (Figure 4) if chromogranin A were indeed an integral membrane protein (Settleman et al., 1985a). We looked for possible sites for protein modifications in chromogranin A. As shown in Figure 2B, chromogranin A contains eight potential dibasic cleavage sites (see Docherty and Steiner, 1982), six of which are located in the C-terminal domain. Chromogranin A does not contain a consensus sequence for N-glycosylation (NXS or NXT, see Kornfeld and Kornfeld, 1985), which agrees with biochemical data showing that glycosylation of the protein is of the 0-linked type (Kiang et al., 1982; Apps et al., 1985). None of the four tyrosine residues is preceded by acidic amino acid residues, which are typically found in this position in the known tyrosine sulfation sites of secretory proteins (see Huttner et al., 1986; Lee and Huttner, 1985). The absence of acidic amino acid residues on the N-terminal side of tyorisine residues may explain why chromogranin A, in contrast to secretogranins I and II, is not tyrosine-sulfated (Rosa et al., 1985b). A comparison of the chromogranin A sequence with the protein sequence data library of the National Biomedical Research Foundation (NBRF) and of the EMBL showed no significant homologies with sequenced proteins. However, we have noted that positions 43-45 of the chromogranin A sequence contain -Arg-Gly-Asp- (RGD). This three-amino-acid sequence has been shown to be involved in the receptor-mediated binding of several secretory proteins, notably fibronectin, to cell membranes (see Ruoslahti and Pierschbacher, 1986).
Fig. 6. Northern blot analysis of chromogranin A. Poly(A)+-RNA (5 ,g per lane) from bovine adrenal, purified by oligo-dT-cellulose chromatography, was hybridized with Klenow-extended single-stranded probes from M13-subclones (see Figure 2A), corresponding to the N-terminal (lane 1), middle (lane 2) and C-terminal (lane 3) portions of the coding region of the chromogranin A mRNA. The positions of the 18S and 28S rRNAs are indicated.
The chromogranin A mRNA Northern blot analysis of poly(A) +-RNA from bovine adrenals with probes corresponding to three different portions of the coding region of the chromogranin A mRNA (compare Figure 2A) indicated that the chromogranin A mRNA has an apparent length of 2.1 kb (Figure 6). Under stringent -conditions, no crosshybridization of any of the three probes to mRNAs of different lengths was observed, suggesting that the mRNAs for secretogranins I and II were not recognized under the present experimental conditions. In vitro transcription -translation of the chromogranin A cDNA The Narl fragment of clone CgA1, containing the entire coding region for the chromogranin A precursor, was subcloned into the expression vector pDS6. The construct was transcribed in vitro, and the resulting capped mRNA was translated in a wheat germ cell-free system. The polypeptide synthesized was heatstable, like chromogranin A (Rosa et al., 1985b), and migrated similarly to authentic chromogranin A upon SDS-PAGE (Figure 7). [The slightly slower electrophoretic mobility of chromogranin A from chromaffin granules compared to the in vitro transcribed-translated chromogranin A precursor is most likely due to post-translational modifications (see Falkensammer et al., 1985a; Rosa et al., 1985b).] The protein product encoded by pDS6-CgAl had an apparent molecular weight of 72 000, which markedly differs from the predicted molecular mass of 50 kd for
1499
U.M.Benedum et al.
-97
-n67
-45
------cye
AB
Fig. 7. In vitro transcription-translation of the chromogranin A precursor. The plasmids pDS6 (lane A) and pDS6-CgA1 (lane B) were transcribed in vitro, and the resulting capped mRNAs translated in a wheat germ cell-free system using [35S]methionine. The translation products were boiled, and the heat-stable proteins separated on a 7.5% SDS polyacrylamide gel. A fluorogram of the gel is shown. The positions of molecular weight markers, detected by Coomassie Blue staining, are indicated. The bracket indicates the position of unlabeled mature (0-glycosylated, phosphorylated, sulfated) chromogranin A from bovine chromaffin granules, which is known to display heterogeneity in SDS polyacrylamide gels (apparent mol. wt 72 000-77 000). The major product encoded by pDS6-CgAl has an apparent mol. wt. of 72 000.
the chromogranin A precursor. This indicates that chromogranin A has a highly abnormal electrophoretic mobility in SDS
polyacrylamide gels. Discussion The primary structure of chromogranin A reported here has significant implications as to the still unknown function of this widespread secretory protein. Three aspects of the structure, discussed below, deserve particular attention: the repeated occurrence of clusters of glutamic acid residues, the presence of several potential dibasic cleavage sites, and the RGD sequence. The structure of chromogranin A The conclusion that the present cDNA clone contains the entire coding region for chromogranin A is based on the agreement of the nucleotide sequence with N-terminal, internal and Cterminal amino acid sequence data as well as the amino acid composition of the protein. Further evidence for the completeness of the cDNA clone comes from the observations that (i) its length was not significantly different from that of the hybridizing mRNA of bovine adrenal and (ii) the product of coupled in vitro transcription-translation migrated similarly to authentic chromo1500
granin A upon SDS-PAGE. The deduced molecular weight of the unmodified polypeptide chain of chromogranin A after removal of the signal peptide is 48 132. This value agrees well with the molecular weight of 53 000 for the mature protein (posttranslationally modified by addition of carbohydrate, phosphate and sulfate), as determined by its sedimentation equilibrium in 6 M guanidine hydrochloride (Kirschner, 1974). However, the deduced molecular weight is considerably less than that of most other previous estimates which were based on analytical ultracentrifugation in the absence of guanidine hydrochloride and on SDSPAGE of the mature protein (apparent mol. wt. 70 000-80 000; see Winkler, 1976; Winkler and Carmichael, 1982). As to the latter method, the discrepancy is explained by our results of in vitro transcription-translation showing that chromogranin A has a highly abnormal mobility upon SDS-PAGE. The present results do not support previous data (Settleman et al., 1985b) on amino acid sequence homologies or gene duplication events within the chromogranin A molecule. The sequences of the 30-kd, 15-kd and 12-kd cyanogen bromide fragments of chromogranin A determined by Settleman et al. (1985b) are all best aligned to only one region of the present sequence (residue 285 to residue 308), although there are several mismatches. Moreover, we do not find any homologies to the Nterminal sequence of intact chromogranin A anywhere else in the molecule. Rather, the N-terminal sequence of the 65-kd fragment of chromogranin A reported by Settleman et al. (1985b) corresponds to the N-terminus of intact chromogranin A, the five amino acid differences (see Figure 1) possibly arising from variability in the animals from which chromogranin A was purified or from errors in protein sequencing. A structural feature that is repeated in the chromogranin A molecule are clusters of glutamic acid residues. We have previously suggested that chromogranin A as well as secretogranins I and II, by means of a high density of negative charges, might function as helper proteins in the packaging of regulatory peptides into secretory granules (Rosa et al., 1985a, b). This hypothesis can now be tested for chromogranin A by using molecular genetics. Is chromogranin A a precursor for regulatory peptides? The N-terminal sequence of p62 (Rosa et al., 1985b; see footnote), a proteolytic product of chromogranin A found in chromaffin granules, is identical to that of chromogranin A (data not shown). We therefore conclude that proteolytic fragmentation of chromogranin A begins in its C-terminal domain. It is worth noting that this domain contains six sites with two adjacent basic amino acids. Cleavage of regulatory peptides from their precursors usually starts at dibasic sites and is thought to occur in secretory granules (see Docherty and Steiner, 1982). Since chromogranin A is stored in secretory granules, it would be exposed to such processing enzymes, and proteolytic fragmentation of chromogranin A may well take place as the result of cleavage at the dibasic sites in the C-terminal domain. It will be interesting to investigate the possibility that chromogranin A may be a precursor of regulatory peptides. This could be tested by analyzing the biological activities of peptides synthesized according to the predictions of the chromogranin A sequence. Possible roles for the RGD sequence in chromogranin A The RGD sequence, first recognized as the cell attachment site in fibronectin, has recently received considerable attention (see Ruoslahti and Pierschbacher, 1986). A computer search in the NBRF data library reveals the presence of this sequence in about 15 different eucaryotic secretory proteins. In at least one-third
Molecular cloning of chromogranin A
of these proteins (fibrinogen, fibronectin, vitronectin, von Willebrand factor and discoidin I of Dictyostelium), where the RGD sequence has been tested for its physiological significance, it has been implicated in the binding of these proteins to cell membranes. At least three different cell surface receptors which recognize the RGD sequences in the various secretory proteins have been identified so far (see Ruoslahti and Pierschbacher, 1986). According to the secondary structure predictions, the RGD sequence in chromogranin A is exposed at the surface of the molecule (data not shown). The presence of the RGD sequence in chromogranin A may therefore have significant functional implications. Firstly, the RGD sequence may mediate the interaction of chromogranin A after secretion with cell surface receptors for other, adhesive proteins containing this sequence such as fibronectin or vitronectin. Secondly, considering the fact that chromogranin A is released with a variety of regulatory peptides from secretory granules, it is attractive to think that the RGD sequence may mediate the binding of chromogranin A to target cells where chromogranin A might modulate the response of the target cell to the regulatory peptide. Thirdly, the RGD sequence may also have a role in the intracellular transport of chromogranin A by mediating the binding of this protein to endomembranes such as those of the trans Golgi network (see Griffiths and Simons, 1986) where proteins like chromogranin A, destined for the regulated pathway of secretion, are sorted. Clearly, the presence of the RGD sequence in a regulated secretory protein with a widespread distribution like chromogranin A deserves further attention. Materials and methods Sequencing of chromogranin A and of chromogranin A fragments Tryptic peptides of chromogranin A. One mg of protein of the soluble fraction of bovine chromaffin granules (Barlett and Smith, 1974; Rosa et al., 1985b) was subjected to SDS-PAGE on a 10% gel. Proteins were stained for 10 min with Coomassie Blue and destained as described (Lee and Huttner, 1983). The central portion of the protein band corresponding to chromogranin A was excised (see Figure 3). After several washes in distilled water, the gel piece (-2 ml of wet gel) was minced, lyophilized, re-swollen in 10 ml of a solution containing 50 mM ammonium bicarbonate, 0.1 mM CaCl2 and 20 yg TPCK-trypsin (Worthington), and incubated for 24 h at 30°C. The eluate was then separated from the gel and incubated for a further 14 h after addition of another 10 yg of TPCKtrypsin, while the gel was re-eluted twice with 10 ml and 2.5 ml each of fresh buffer containing a total of 10 Ag of TPCK-trypsin. The eluates were pooled and passed several times through a C18-cartridge (SEP-PAK, Waters) until all the Coomassie Blue had bound. Peptides were eluted stepwise with 30, 50 and 100% methanol, and the fractions were concentrated in a Speedvac, adjusted to 20% (v/v) formic acid, and analyzed by h.p.l.c. as described in the legend to Figure 3. We used 99.9% (v/v) water/0.1% (v/v) TFA and 99.93% (v/v) acetonitrile/0.07% (v/v) TFA as the two stock solutions for the gradients. Individual peaks were collected, concentrated in a Speedvac, and analyzed by Nterminal sequencing in a gas-phase sequenator as described (Frank and Trosin, 1985). N-terminus and C-terminus of chromogranin A. After SDS-PAGE as above, the gel was stained with ice-cold 0.4 M potassium acetate, the band corresponding to chromogranin A was excised, minced, and chromogranin A electro-eluted as described (Knowles and Bologna, 1983) except that BICIN (Merck) was used instead of glycin. Electro-eluted chromogranin A (300 Al) was diluted into 18 ml methanol, cooled to - 70°C, and the precipitate was collected by centrifugation for 10 min at 10 000 g. For N-terminal sequencing, the pellet was dissolved in 1% (v/v) TFA and analyzed in a gas-phase sequenator as described (Frank and Trosin, 1985). For C-terminal sequencing, the pellet was dissolved in 300 tlI of 0.1% (v/v) pyridine adjusted with dilute acetic acid to pH 6.5. The solution was warmed to 37°C, 3 jig each of carboxypeptidase B and P (Boehringer) dissolved in the pyridine-acetate buffer were added, and the sample was incubated at 37°C. Immediately after the addition of the enzymes and at the times indicated in Figure 4, 40 itl aliquots were removed and added to 10 1l of 0.1% (w/v) iodoacetic acid to stop the carboxypeptidase digestion. Aliquots taken from a reaction containing carboxypeptidases, the pyridine-acetate buffer, but no electro-eluted chromogranin A served as controls for autodigestion of the enzymes. All samples were subjected to amino acid anlaysis using the orthophthaldialdehyde method as described (Ashman and Bosserhoff, 1985).
Cloning of chromogranin A cDNA The poly(A)+-RNA from bovine adrenal medulla was purified according to Lomedico and Saunders (1976). The cDNA library was constructed in pBR322 according to the method of Gubler and Hoffman (1983). The 59-mer oligonucleotide used for screening the library was produced with an Applied Biosystems 380A synthesizer by the phosphoramidite-method and purified by PAGE. One hundred nanogrammes of the oligonculeotide were labelled with ['y32P]ATP (NEN, 7000 Ci/mmol) using T4-polynucleotide kinase (Boehringer), which resulted in complete conversion of the label (Lillehaug et al., 1976). The labeled oligonucleotide was purified on a sequencing gel. A total of approximately 14 000 different colonies were screened. Seven thousand colonies were transferred to each 20 x 20 cm nitrocellulose filter and lysed according to Grunstein and Hogness (1975), and the filters were prehybridized at 37°C for 4 h in hybridization buffer [6 x SSC; 5 x Denhardt's (1966); 50 mM sodium phosphate, pH 6.5; 100 jig/ml boiled herring DNA; 20% (v/v) deionized formamide; 0.1 g/ml dextran sulfate, mol. wt 500 000]. The filters were then hybridized at 38-40°C for 12 h with the oligonucleotide (150 000 c.p.m./ml) in hybridization buffer. The nitrocellulose filters were washed at 37°C for 30 min in 1 x SSC/0. 1% SDS with several changes of buffer and exposed overnight with an intensifying screen at -70°C. Positive colonies were picked and rescreened under the same conditions. DNA and RNA blot hybridization and primer extension DNA blot hybridizations were carried out according to the standard method (Southern, 1975). Total cellular RNA was isolated from bovine adrenal according to the method of Chirgwin et al. (1979). Poly(A)+-RNA was purified by two cycles of oligo-d(T)-cellulose column chromatography and recovered by ethanol precipitation. mRNA to be analysed by the method of Thomas (1980) was denatured at 70°C in the presence of 50% formamide for 10 min prior to electrophoresis on Tris/borate/EDTA agarose gels in the presence of 0.5 tg/ml ethidium bromide. Nitrocellulose blots (BA85, Schleicher and Schuell) were prehybridized for 5 h at 42°C in 50 mM sodium phosphate, pH 7.0, 5 x SSC, 50% formamide, 5 x Denhardt's, 0.0125% SDS and 100 jLg/ml of boiled and sonicated herring sperm DNA. Hybridizations were performed for 16 h at 420C in 50 mM sodium phosphate, pH 7.0, 5 x SSC, 45% formamide, x Denhardt's, 0.4 % SDS, 100 Ag/ml of boiled and sonicated herring DNA and 1.6 x 106 c.p.m. of the respective single-stranded probe. Specific singlestranded probes were generated in the course of DNA-sequencing (see Figure 2A). They were labeled with [a 32P]dATP and [ca32P]dCTP (3000 Ci/mmol) according to Messing (1983). The primer-extended M13 probes were then cut with PstI, denatured by boiling for 3 min in 80% deionized formamide, and separated on a Tris/borate/EDTA low melting agarose gel (Burke, 1984). The radioactive single-stranded fragments were isolated from the gel (Maniatis et al., 1982) and used as hybridization probes. Sequencing of chromogranin A cDNA A 2.1-kb ScaI fragment which contained the entire cDNA insert except an 100-bp fragment at the 3'-end (see Figure 2A) was isolated and subcloned in both orientations into M13mpl8 and M13mpl9. DNase I deletions were generated according to the method of Frischauf et al. (1980), using EcoRI and Hindlll for M13mpl8 and M13mpl9 clones (Messing, 1983), respectively. Blunt ends were produced with DNA polymerase I (Klenow fragment) and religated (0.5 'ag DNA/ml) with T4-ligase. To reduce background after transformation, the M13 was cut with SalI or KpnI (Lin et al., 1985). DNA sequence determination was carried out beginning with the largest insert and proceeding to the smaller ones so that overlapping sequence data were obtained. The portion of the sequence not covered by the deletion cloning was determined from M13 subclones containing PstI fragments. The M13 subclones obtained were subjected to DNA sequence analysis according to the Sanger method (Sanger et al., 1977), incorporating the modifications reported by Biggin et al. (1983). In vitro transcription -translation of cloned chromogranin A Coupled transcription -translation was performed as described by Stueber et al. (1984) utilizing the plasmid vector pDS6. The recombinant employed, designated pDS6-CgAl, was constructed in the following manner. Purified plasmid DNA from clone CgA1 was digested with Narn, and the 1367-bp coding fragment (see Figure 2A) isolated from low-melting agarose after electrophoresis. This fragment was rendered blunt-ended using DNA polymerase I (Klenow fragment) and subcloned into dephosphorylated SinaI-cut M13mp8 RF DNA. The M13 subclones were analyzed with regard to the orientation of the inserts. The double-stranded DNA from an M13 subclone with the correct orientation was restricted with EcoRI and HinduI. The resulting insert was purified by low-melting agarose gel electrophoresis and ligated into pDS6 linearized by digestion with EcoRI and HindIll. The resulting subclone pDS6-CgA1 was propagated in LB medium in the presence of 50 jLg/ml ampicillin and DNA isolated according to the rapid alkaline extraction procedure of Birnboim and Doly (1979). This DNA was further purified after linearization with HindIlI by low-melting agarose gel electrophoresis prior to in vitro transcription -translation (Stueber et al., 1984). Following translation
1501
U.M.Benedum et al. using [35S]methionine, the reactions were supplemented with the soluble protein fraction from bovine chromaffin granules (5 jg chromogranin A) and boiled for 3 min in the presence of 100 mM NaCl. Denatured proteins were removed by centrifugation for 15 min at 13 000 g. The synthesized [35S]methionine-labeled, heat-stable proteins were analyzed by SDS-PAGE (Laemmli, 1970) on 7.5% gels and visualized by fluorography. Miscellaneous For the comparative analysis of sequence data, the computer programme of Lipman and Pearson (1985) was used. The hydropathy plot was calculated according to Kyte and Doolittle (1982). Footnote on the nomenclature To avoid confusion, the term 'chromogranin B' will not be used in this study. It has been used by O'Connor and Frigon (1984) to refer to a proteolytic fragment of chromogranin A, called the 65-kd polypeptide by Settleman et al. (1985b) and p62 by Rosa et al. (1985b). More recently, this term has also been used by Winkler and colleagues (e.g., Falkensammer et al., 1985a, b) to refer to secretogranin I (Lee and Huttner, 1983; Rosa et al., 1985b). Also, for ease of presentation we shall use the term chromogranin A when referring to this protein in the parathyroid, where it was originally called secretory protein I (see Cohn and MacGregor, 1981).
Acknowledgements We are grateful to Professor H.Thoenen for his support during the initial part of this study at the Max-Planck-Institute for Psychiatry and to Professor E.Fanning for her support of the doctoral thesis of U.B. We thank Drs A.Lamouroux and J.F.Julien for the cDNA library and help, Dr H.J.Fritz of the Max-PlanckInstitute for Biochemistry for synthesizing the 59-mer oligonucleotide, A.Bosserhoff and H.Gausepohl for help with amino acid analysis and sequencing, Dr P.Rosa for PC12 cells, chromogranin A, anti-chromogranin A antibody and advice, Dr S.Saadat and U.Seydel for purified chromaffin granules, Drs B.Dobberstein and M.-T.Haeuptle for advice concerning coupled transcription-translation, Drs P.Argos, D.Banner and R.Heumann for helpful discussions, and Drs R.Cortese, S.Fuller, A.Hille, R.W.H.Lee, D.I.Meyer, P.Rosa and K.Stanley for their helpful comments on the manuscript. J.M. was supported by the CNRS and W.B.H. by grants from the Deutsche Forschungsgemeinschaft (Hu 275/3-2, Hu 275/3-3).
References Apps,D.K., Phillips,J.H. and Purves,F.C. (1985) Neuroscience, 16, 477-487. Ashman,K. and Bosserhof,A. (1985) In Tschesche,H. (ed.), Modem Methods in Protein Chemistry. de Gruyter, Berlin and New York, pp. 155-171. Bartlett,S.F. and Smith,A.D. (1974) Methods Enzymol., 31, 379-394. Biggin,M.D., Gibson,T.J. and Hong,G.F. (1983) Proc. Natl. Acad. Sci. USA, 80, 3963-3965. Birnboim,H.C. and Doly,J. (1979) Nucleic Acids Res., 7, 1513-1523. Burke,J.F. (1984) Gene, 30, 63-68. Chirgwin,J.M., Przybyla,A.E., MacDonald,R.J. and Rutter,W.J. (1979) Biochemistry, 18, 5294-5299. Cohn,D.V. and MacGregor,R.R. (1981) Endocrinol. Rev., 2, 1-26. Cohn,D.V., Morrissey,J.J., Hamilton,J.W., Shofstall,R.E., Smardo,F.L. and Chu,L.L.H. (1981) Biochemistry, 20, 4135-4140. Cohn,D.V., Zangerle,R., Fischer-Colbrie,R., Chu,L.L.H., Elting,J.J., Hamilton,J.W. and Winkler,H. (1982) Proc. Natl. Acad. Sci. USA, 79, 6056-6059. Cohn,D.V., Elting,J.J., Frick,M. and Elde,R. (1984) Endocrinology, 114, 1963-1974. Denhardt,D.T. (1966) Biochem. Biophys. Res. Commun., 23, 641-646. Docherty,K. and Steiner,D.F. (1982) Annu. Rev. Physiol., 44, 625-638. Falkensammer,G., Fischer-Colbrie,R., Richter,K. and Winkler,H. (1985a) Neuroscience, 14, 735-746. Falkensammer,G., Fischer-Colbrie,R. and Winkler,H. (1985b) J. Neurochem., 45, 1457-1480. Fischer-Colbrie,R., Lassmann,H., Hagn,C. and Winkler,H. (1985) Neuroscience, 16, 547-555. Frank,R. and Trosin,M. (1985) In Tschesche,H. (ed.) Modem Methods in Protein Chemistry. de Gruyter, Berlin and New York, pp. 287-302. Frischauf,A.M., Garoff,H. and Lehrach,H. (1980) Nucleic Acids Res., 8, 5541 -5549. Grantham,R., Gautier,C., Guoy,M., Mercier,R. and Pave,A. (1980) Nucleic Acids Res., 8, 49-62. Griffiths,G. and Simons,K. (1986) Science, in press. Grunstein,M. and Hogness,D. (1975) Proc. Natl. Acad. Sci. USA, 72, 3961 -3965. Gubler,U. and Hoffman,B.J. (1983) Gene, 25, 263-269. Hogue Angeletti,R. (1977) Arch. Biochem. Biophys., 184, 364-372.
1502
Hortsch,M. and Meyer,D.I. (1986) Int. Rev. Cytol., 102, 215-242. Huttner,W.B., Baeuerle,P.A., Benedum,U.M., Friederich,E., Hille,A., Lee,R.W.H., Rosa,P., Seydel,U. and Suchanek,C. (1986) Hormones and Cell Regulation, Vol. 10, in press. Kelly,R.B. (1985) Science, 230, 25-32. Kiang,W.-L., Krusius,T., Finne,J., Margolis,R.U. and Margolis,R.K. (1982) J. Biol. Chem., 275, 1651-1659. Kirshner,N. (1974) In Cecarelli,B., Clementi,F. and Meldolesi,J. (eds), Advances in Cytopharmacology. Vol. 2, pp. 265-272. Knowles,W.J. and Bologna,M.L. (1983) Methods Enzymol., 96, 305-313. Kornfeld,R. and Kornfeld,S. (1985) Annu. Rev. Biochem., 54, 631-664. Kruggel,W., O'Connor,D.T. and Lewis,R.V. (1985) Biochem. Biophys. Res. Commun., 127, 380-383. Kyte,J. and Doolittle,R.F. (1982) J. Mol. Biol., 157, 105-132. Laemmli,U.K. (1970) Nature, 227, 680-685. Lee,R.W.H. and Huttner,W.B. (1983) J. Biol. Chem., 258, 11326-11334. Lee,R.W.H. and Huttner,W.B. (1985) Proc. Natl. Acad. Sci. USA, 82, 6143-6147. Lillehaug,R.E. Kleppe,R.K. and Kleppe,K. (1976) Biochemistry, 15, 1858-1865. Lin,H., Lei,S. and Wilcox,G. (1985) Anal. Biochem., 147, 114-119. Lipman,D.J. and Pearson,W.R. (1985) Science, 227, 1435-1441. Lloyd,R.V. and Wilson,B.S. (1983) Science, 222, 628-630. Lomedico,P.T. and Saunders,G.F. (1976) Nucleic Acids Res., 3, 381-391. Maniatis,T., Fritsch,E.F. and Sambrook,J. (1982) Molecular Cloning, A Laboratory Manual. Cold Spring Harbor Publications, Cold Spring Harbor, NY. Messing,J. (1983) Methods Enzymol., 101, 20-78. O'Connor,D.T., Burton,D. and Deftos,L.J. (1983) Life Sci., 33, 1657-1663. O'Connor,D.T. and Frigon,R.P. (1984) J. Biol. Chem., 259, 3237-3247. Rosa,P. and Zanini,A. (1981) Mol. Cell. Endocrinol., 24, 181-193. Rosa,P. and Zanini,A. (1983) Eur. J. Cell Biol., 31, 94-98. Rosa,P., Fumagali,G., Zanini,A. and Huttner,W.B. (1985a) J. Cell Biol., 100, 928-937. Rosa,P., Hille,A., Lee,R.W.H., Zanini,A., De Camilli,P. and Huttner,W.B.
(1985b) J. Cell Biol., 101, 1999-2011. Ruoslahti,E. and Pierschbacher,M.D. (1986) Cell, 44, 517-518. Sanger,F., Nicklen,S. and Coulson,A.R. (1977) Proc. Natl. Acad. Sci. USA,
74, 5463-5467. Settleman,J., Nolan,J. and Hogue Angeletti,R. (1985a) J. Biol. Chem., 260, 1641-1644. Settleman,J., Fonseca,R., Nolan,J. and Hogue Angeletti,R. (1985b) J. Biol. Chem., 260, 1645-1651. Somogyi,P., Hodgson,A.J., DePotter,R.W., Fischer-Colbrie,R., Schober,M., Winkler,H. and Chubb,I.W. (1984) Brain Res. Rev., 8, 193-230. Southern,E. (1975) J. Mol. Biol., 98, 503 -505. Stueber,D., Ibrahimi,I., Cutler,D., Dobberstein,B. and Bujard,H. (1984) EMBO
J., 3, 3143-3148. Thomas,P.S. (1980) Proc. Natl. Acad. Sci. USA, 77, 5201-5205. Winkler,H. (1976) Neuroscience, 1, 65-80. Winkler,H. and Carmichael,S.W. (1982) In Poisner,A.M. and Trifaro,J.M. (eds), The Secretory Granule. Elsevier Biomedical Press, pp. 3-78. Zanni,A. and Rosa,P. (1981) Mol. Cell. Endocrinol., 24, 165-179. Received on 18 April 1986; revised on 7 May 1986
